节点文献

高速机动目标雷达信号参数估计与成像处理

Radar Signal Parameter Estimation and Image Processing of High-speed Maneuvering Target

【作者】 李文臣

【导师】 王国玉;

【作者基本信息】 国防科学技术大学 , 信息与通信工程, 2009, 博士

【摘要】 随着国内外导弹突防和反突防技术的发展,高速机动已成为目标突防的主要手段之一。高速机动目标具有大的雷达径向速度和加速度,目标高速机动将使防御雷达常规信号处理的分辨性能和信噪比下降,宽带一维距离像展宽,二维像模糊,因此将最终导致防御雷达无法对高速机动突防目标进行探测、跟踪与成像。如何提高防御雷达对高速机动目标的探测、跟踪和成像能力已成为雷达领域的前沿课题和紧迫任务。以弹道导弹防御为研究背景,论文围绕高速机动目标雷达信号参数估计与成像处理,深入系统地研究了高速机动目标的雷达信号模型、多分量LFM信号调频参数估计方法、基于二次匹配处理的多分量LFM信号分辨力、以及高速机动目标宽带一维像补偿和二维成像及定标。论文研究成果具体体现在以下几个方面:1.提取了高速机动目标在常规窄带雷达和宽带成像雷达下的回波特征。为了提高高速机动目标的雷达探测性能,需要研究高速机动目标的运动特征及回波模型,以便进一步研究雷达信号处理或补偿方法。为此首先分析了高速机动对雷达径向速度和加速度带来的影响。其次提取了高速机动目标在常规窄带雷达和宽带成像雷达下的回波特征,分析了常规窄带雷达信号处理方法的局限性。提出了用展宽系数来衡量频谱展宽的方法,并定量分析了目标运动各参数对宽带雷达测距和一维像的影响。最后分析了多分量LFM信号的时频分布和模糊图特征,以便选择合理的参数估计方法。2.研究了多分量LFM信号调频参数估计方法。高速机动目标的雷达回波信号处理均和LFM信号的检测和参数估计有关。为了研究LFM信号的二次匹配处理与参数估计,首次从理论上系统研究了LFM信号Radon-Wigner变换(RWT)、Wigner- Hough变换(WHT)、分数阶傅立叶变换(FrFT)、解线调方法(Dechirp)、Chirp-Fourier变换(CFT)和最大似然(ML)等线性调频参数估计方法的相互转换关系,给出了LFM信号调频参数估计的最佳表述。首次系统地给出了基于RWT的LFM信号快速调频参数估计,提出了利用频率偏差系数和剩余展宽系数分析LFM信号参数估计性能的方法。并提出了基于Radon-Ambiguity变换(RAT)的LFM信号快速调频参数估计。计算量比较分析表明RAT和RWT方法的运算量是相近的,估计性能均能达到了Cramer-Rao门限(CRB)。RWT法能同时估计LFM信号的频率和调频率,适合分析任意多分量LFM信号,由于能够充分利用各分量信号的能量,因此RAT适合估计同调频率多分量信号的调频率。3.基于RWT方法研究了多分量LFM信号分辨力和多目标分辨力。首次用二次项泰勒级数展开给出了RWT输出的近似解析表达式,从理论上给出了频率和调频率联合分辨数学表达式,以及延迟时间、频率和调频率的联合分辨力,并得到仿真验证。然后给出了脉冲多普勒雷达的速度和加速度联合分辨表达式,表明其联合分辨条件为运动参数不同引起的路径偏差大于半个波长,仿真比较了傅立叶变换和RWT的性能。最后给出了LFM雷达的距离、速度和加速度的联合分辨表达式,表明其分辨条件决定于两目标的径向位置偏差(分辨单元级)和相参时间内径向运动距离偏差(波长级),并仿真比较了匹配滤波和RWT的性能。本研究有助于机动目标的跟踪,以及群目标的分辨与识别。4.系统研究了高速机动目标的宽带雷达运动补偿、目标连续跟踪测量和ISAR成像与定标。根据速度和加速度的调频频谱展宽特点,提出了机动目标宽带一维距离像回波线性化模型,给出了RAT法线性参数估计与运动补偿方法,并进一步分析了测速和测距误差。仿真表明该方法很好地解决了未知运动参数情况下机动目标的一维距离像频谱展宽问题。其次为了有效利用雷达资源,一旦窄带跟踪到目标后,雷达可以转入宽带一维成像与单脉冲跟踪。给出了宽带一维像的测速、测距和测角模型,并得到仿真验证。目标一维像的连续定位跟踪可以增加目标跟踪数据率,提高雷达对机动目标跟踪能力。提出了匀加速目标的变尺度RWT成像与定标方法。该方法首先根据多个距离单元信号的调频参数估计,用加权最小二乘法解算出目标转动参数,并利用转动参数对横向信号实施纵向参数补偿;最后利用变尺度RWT实现成像与定标。算法速度较常规RWT成像有大幅度提高,成像性能更稳定,仿真实验和外场数据的ISAR成像处理验证了该方法的有效性。该方法在非合作机动目标ISAR成像与定标方面有很好的应用前景。另外纵向参数补偿法对于匀速运动目标仍然具有很好的应用价值。另外为了验证变尺度RWT成像,提出了基于外场测试数据的插值变采样机动目标数据产生方法,并进行了仿真验证。最后基于对称目标的对称轴与对称点直线垂直,给出了对称目标转角估计和横向定标新方法,并给出了定标参数估计误差表达式,定量分析了影响定标参数估计误差的各因素,并进行了仿真验证。该方法结构简单,运算量小。论文研究成果可以应用于脉冲多普勒雷达、LFM连续波雷达、步进频雷达、远程预警雷达和成像雷达等,对于雷达信息处理等诸多领域均提供借鉴和参考,同时对于弹道导弹生存和突防能力等应用领域研究具有十分重大的理论意义和军事价值。

【Abstract】 Along with the unceasing development of domestic and foreign missile defense penetration and counter-defense penetration technology, the high-speed maneuver has become the main means of defense penetration. The high-speed maneuvering target is provided with a big radar radial velocity and acceleration, if the conventional signal processing method is used, the resolution performance and signal noise ratio of defense radar decrease, and the wideband range profile is broadening, and the two-dimension image is blurring, which induce that the defense radar fail in detection, tracking and image formation for the high-speed maneuvering target. How to enhance the radar in target detection, tracking and image formation for the high-speed maneuvering target already became the front topic and the pressing task for the current radar technology domain.This dissertation takes the ballistic missile defense as the research background, focuses on radar signal parameter estimation and image processing of the high-speed maneuvering target,and systematically studies the radar signal model, and the related processing method of the high-speed maneuvering target, including frequency modulation (FM) parameters estimation of multi-component LFM signals, the resolution of multi-component LFM signals based on quadratically matched processing, motion compensation of wideband range profile, two-dimension imaging and cross-range calibration. The research results mainly involve:1) The high-speed maneuvering target’s echo characteristics of both conventional narrow band radar and wide band imaging radar are extracted. In order to improve the radar exploration performance of the high-speed maneuvering target, the target’s motion characteristics and echo signal model, and the radar signal processing and compensating methods should be studied. Firstly, the radial velocity and acceleration influence brought by the high-speed maneuver are analyzed. Then the echo signal models of narrow-band radar and wideband imaging radar are extracted, and the limitations of conventional processing method are analyzed. Spectral expansion coefficient is proposed to measure spectral expansion, and range measuring and spectral expansion effects caused by movement parameter are analyzed quantificational. Finally, in order to select the optimal FM parameters estimation method, the time-frequency distribution and the ambiguity chart characteristic have been analyzed for the multi-component LFM signals.2) The FM parameters estimated methods of the multi-component LFM signals have been studied. The radar echo processing of the high-speed maneuvering target is related with detection and parameter estimation of the LFM signals. In order to study the quadratically matched processing and parameter estimation of the LFM signals, the reciprocal transforming relationship of FM parameters estimated methods for the continuous complex LFM signals, including Radon-Wigner transform (RWT), Wigner-Hough transform (WHT), Fractional Fourier transform (FrFT), Dechirp,Chirp-Fourier transform (CFT) and Maximum Likelihood (ML), are studied in theory for the first time, and the optimal expression of FM parameters estimation is proposed. Secondly, Multi-resolution fast parameters estimation arithmetic of LFM signals via RWT is offered. The parameters estimation performance is studied on frequency deviation coefficient and residual expansion coefficient, and the computational cost is analyzed. Finally, the fast algorithm of Radon-Ambiguity transform (RAT) method of the discrete signals is proposed. Simulation results have validated that both RWT and RAT can reach the Cramer-Rao bound (CRB) even on low SNR. The analysis indicates that RWT accesses RAT in computational cost. The RWT method can estimate the frequency and chirp rate of the LFM signals, and adapt to analyze arbitrary the multi-component LFM signals. The RAT method fits to estimate the chirp rate of the multi-component LFM signals in the same chirp rate, because the total energy of various component signals can be utilized.3) Based on RWT method, the resolutions of multi-component LFM signals and multi-target are studied respectively. The RWT output approximate expression is given based on quadratic Taylor’s series expansion for the first time, and the joint resolution mathematical expression of the FM parameters, and the joint resolution of delay time, frequency and chirp rate are offered in theory, which are validated by simulation results. Then, the joint resolution expression of velocity and acceleration of the pulse Doppler radar is offered. It is indicated that the motion parameters can be distinguished if the radial distance deviation reaches half wave length. The simulation results have been given to compare the performance of the Fourier transform and the RWT method. Finally, the joint resolution of distance, velocity and acceleration for LFM radar is studied, which indicated that the joint resolution condition is decided by the radial position deviation (resolution cell level) and the radial distance moving deviation in coherent time (wave length level) of the two targets. The simulation results have been given to compare the performance of the matched filtering and the RWT method. This research is helpful to track maneuvering target, as well as group target resolution and recognition.4) The motion compensation, continuous tracking, ISAR imaging and cross-range calibration of the wideband radar for the high-speed maneuvering target are studied systemically. Based on the spectral expansion characteristics of the velocity and acceleration, a novel linear model of wideband range profile of maneuvering target is proposed. Linear parameter estimation and compensation of range profile via RAT are presented, and the measurement errors of range and velocity are analyzed. The simulation results show that this method can resolve the range profile spectral expansion of maneuvering target unknown motion parameters. In order to use radar resource effectively, once the target has been tracked by narrow band signal, the radar can switch to mono-pulse tracking and imaging of wideband signal. Angle, distance and velocity measurement models of wideband range profile are presented, which are validated by simulation results. Target orientation and tracking based on range profile can increase data rate of target tracking, which can enhance the tracking capability of maneuvering target.The scaled RWT imaging and cross-range scaling method is proposed for the rotating target with stable acceleration. Firstly, the rotating parameters are calculated via weighted least square method after obtaining the FM parameters of signals on multi-range. The slow time signals are compensated using rotating parameters. Then, the ISAR imaging of scaled RWT methods is implemented with cross-range calibration. The calculation speed of this algorithm is improved in great part compared to general RWT imaging, and imaging performance is steady-going. The simulation results and ISAR imaging processing of radar data in outfield validate the method. The method presents good application foreground to the ISAR imaging and calibration of uncooperative motive targets. When target uniform motion, the longitudinal parameter compensation method still have a very good application value.In order to validate scaled RWT imaging method, the ISAR data generating technology of maneuvering target is presented via interpolation sampling based on outfield experimental radar data, and the method is validated by simulation results. Finally, based on the orthogonal relation between the centerline and the line through axially symmetric points in the symmetrical target, a novel method of rotation angle estimation and cross-range calibration is definitely proposed for the symmetrical target. The estimation error expression of the scaling parameter is offered, which is validated by simulation results, and influencing factors of the parameter estimation error is analyzed quantitatively. This method is very simple in structure and have low computation cost.The research results can be applied to pulse Doppler radar, LFM continuous-wave radar, step frequency radar, long-rang early-warning radar and imaging radar, and offer references for many fields such as radar signal processing, and have a great theory significance and military value for the application domain study such as the ballistic missile survival and defense penetration ability.

节点文献中: